Abstract
Background:
Depression and anxiety are common and associated with inflammation in patients with cancer. Inflammatory indices such as albumin and neutrophil-to-lymphocyte ratio (NLR) obtained from metabolic panels and complete blood counts should be available for mental health professionals treating anxiety and depression at cancer centers. We hypothesized that albumin and NLR extrapolated from non-mental health oncology appointments would be associated with anxiety and depression and drawn close enough to psychiatry visits to be useful for the psycho-oncologist.
Materials & Methods:
Depression and anxiety were evaluated in patients (n=97) referred to a cancer center psychiatric service for depression using the Patient Health Questionnaire-9 (PHQ-9) and General Anxiety Disorder-7 (GAD-7). Albumin concentration and NLR were assessed for timing and correlation strength with anxiety and depression by setting (localized/metastatic cancer).
Results:
Most patients (96%) had albumin or NLR available at any timepoint of which 45% were drawn within one week of the psychiatric appointment. No significant correlations were noted when evaluating localized cancer or NLR exclusively. For patients with metastatic cancer, anxiety and depression were correlated with albumin at any timepoint (r=−.28, p<.05; r=−.40, p<.01, respectively) and within a week of psychiatry appointment (r=−.40, p<05; r=−.68, p<.001, respectively). Albumin evaluated within a week predicted 32% of depression score variance (β=−.63, p=.002). Hypoalbuminemia (<3.8 g/ul) was associated with anxiety (χ2=4.43, p=.04) and depression (χ2=11.06, p=.001).
Conclusion:
Hypoalbuminemia in patients with metastatic cancer may help establish the presence or persistence of anxiety, depression, treatment refractoriness, and the use of inflammation in cancer-related psychological symptom management.
Keywords: Psycho-Oncology, depression, anxiety, cancer, albumin, neutrophil-to-lymphocyte ratio, antidepressant medication, cancer-related inflammation
Introduction:
The management of psychiatric symptoms in the cancer context has received increasing attention over the last several decades.(1) Depression and anxiety are highly prevalent across all cancer types and stages and represent one of the most common and serious psychiatric outcomes of cancer.(2) In particular, depression carries the greatest influence on quality of life, adherence with cancer treatments, and adverse consequences on personal relationships and social supports, and is associated with inferior cancer-related survival.(3, 4) Depression with comorbid anxiety confers suicide risk which is already elevated in patients with cancer especially at the beginning of their diagnosis.(5) Although there is significant variability, most cancer centers and oncology practices have deployed some level of psychiatric and palliative or supportive care services that work closely with oncology clinicians to address the highly prevalent psychiatric needs of patients with cancer.(6) These organizational and system level advances offer significant benefits in providing concurrent care, symptom reduction, and enable many cancer centers to achieve a holistic integrated care for their patients.
Importantly, the psychopharmacologic treatment of depression and anxiety has never been biologically targeted for patients with cancer or other chronic medical illness.(7) An inadequate understanding of their driving biological components in the cancer setting or other chronic medical illness contexts has precluded the biological specificity of antidepressant prescribing in clinical practice.(8) The pharmacologic management of depression and anxiety with antidepressants remains a trial and error endeavor while many aspects of cancer care are increasingly personalized.(9) Additionally, while some clinicians may believe that antidepressant therapies work well enough,(10) the rates of depression remission and not just symptom reduction are quite low in non-cancer patients (roughly 30%) and antidepressant medications are generally less effective in patients with chronic medical illness.(11)
Depression is associated with inflammation, (12, 13) and this relationship has been reported in multiple settings.(8) Inflammation is additionally associated with anxiety,(14) though this relationship is less frequently reported in the cancer setting.(15) The high prevalence of depression and anxiety in cancer may stem from cancer-related inflammation, which is considered a hallmark of cancer.(7, 8, 16) The treatment-refractoriness of depression in medical illness may be a reflection, in part, of the impact of inflammation on treatment response, although SSRIs(17) and SNRIs(18) have some minor anti-inflammatory effects.(8) Pro-inflammatory cytokines (i.e., IL-6, TNF) and the positive acute phase reactant C-reactant protein (CRP) have been most extensively studied in association with depression including in the cancer setting and elevation of these inflammatory markers have been shown to predict treatment non-response in ambulatory depressed patients especially to SSRIs and increases in these inflammatory markers are associated with treatment resistance.(19–21) However, these biomarkers are not routinely drawn at medical visits in oncology or other specialties. Standard labs collected at routine visits can be used as indices of inflammation.
Albumin, a negative acute phase reactant whose synthesis by the liver is suppressed by systemic inflammation, has been associated with depression in various medical contexts in addition to its association with nutrition.(22) Similarly, the neutrophil-lymphocyte ratio (NLR) obtained from the white blood cell differential on a complete blood count measure has also been associated with both inflammation and depression in cancer and other clinical settings.(23, 24) We previously studied biomarker changes in patients with metastatic lung cancer and found that albumin changes correlate with changes in CRP and can be used to identify more severe depressive symptoms especially when combined with elevated CRP above 1mg/L.(25) That is, albumin changed with depression severity to a similar extent as CRP and accounted for a similar variability in depression scores regardless of nutritional status (pre-albumin is generally considered a better laboratory marker of nutrition).(26) Albumin has also been associated with depression in several other medical settings (e.g., renal failure, HIV, stroke) independently of nutritional status.(22, 27, 28)
Furthermore, as noted above, inflammation can predict antidepressant non-response (i.e., treatment resistance) and possibly psychopharmacologic treatments for depression that address inflammation are required.(20, 29)
While blood draws are not customary practice for most psychiatry visits, routine blood draws (i.e., from oncology visits) have the potential to direct antidepressant management based on predictive biomarkers of inflammation. However, to mitigate patient burden, especially among those with depression, we speculated that routine laboratory assessments drawn at other cancer-related appointments would be available and provide an indication of the patient’s inflammatory status. Ultimately, this information may be useful to address inflammation in the treatment of depression-as evidence-based treatments become available-and to direct depression care by identifying patients whose depression or anxiety may be refractory to standard non-biologically directed treatments. Specifically, we hypothesized that the negative acute phase reactant, albumin, and the NLR would be associated with depression and anxiety and that the relationship would be strongest for patients with metastatic cancer
Methods and Materials:
Study design
The study was approved by the Memorial Sloan Kettering Cancer Center (MSKCC) Institutional Review Board (IRB) on 12/10/2018 as protocol number 18–526 (Routine inflammatory markers among mental health clinic patients being treated for depression), a retrospective case–cohort. Surveys and blood work were collected over twelve months (May 2017 to May 2018) in a dedicated psychiatric outpatient clinic at a single academic comprehensive cancer center. The data that support the findings of this study are available from the corresponding author, (DM) upon reasonable request.
Participants
Patients who were being treated for depression at a psychiatric service clinic dedicated to the treatment of patients with cancer were included in the study. Inclusion required a confirmed histologic diagnosis of cancer (solid or liquid malignancy) at any stage (metastatic, localized, no evidence of disease). For the purposes of data management and analysis, patients were divided into those with active cancer who may or may not be receiving treatment (i.e., metastatic cancer) versus localized cancer who may be engaged in definitive treatments or undergoing surveillance evaluations.
Procedure
Charts were reviewed retrospectively, and pertinent information was gathered and analyzed. Study consent was granted via the IRB at MSKCC for retrospective data collection.
Measures
Medical and demographic characteristics
Medical information was gathered from the electronic health record. Collected information included patient age, sex, cancer diagnosis and disease types (i.e., solid versus liquid tumors), depression and anxiety scores as continuous and dichotomous variables for meeting screening thresholds, psychiatric notes, and management plans (i.e., was patient taking an antidepressant? Plan to start or increase antidepressant?), timing of closest blood draw (i.e., number of days before or after psychiatric visit), and laboratory values for albumin and Complete Blood Count (CBC).
Albumin
Albumin is a negative acute phase reactant that changes in response to systemic inflammation and was calculated in a Clinical Laboratory Improvement Amendments (CLIA) certified lab.(30) Albumin values were obtained in the same CLIA certified lab.(31) Serum albumin values were obtained by an Albumin BCG assay on the ARCHITECT c Systems used for the quantification of albumin in human serum or plasma. Inter- and intra-assay coefficient of variation is reliably less than 5%.
Neutrophil-to-Lymphocyte Ratio (NLR)
The NLR was obtained from the complete blood count. The ratio was calculated by dividing the Absolute Neutrophil Count (ANC) by the Absolute Lymphocyte Count (ALC). The value was obtained in a Clinical Laboratory Improvement Amendments (CLIA) certified lab.(30) Inter- and intra-assay coefficient of variation is reliably less than 5%.
Depression
Depression severity and screening criteria were measured using the Patient Health Questionnaire-9 (PHQ-9). The PHQ9 is a nine-item measure that scores each question from 0 to 3 for a range of scores from 0 to 27. Depression severity using the PHQ9 has been graded as mild (5–9), moderate (10–14), moderately severe (15–19), and severe (20–27). Depression cutoff scores range from 8–11. Although a lower cutoff score of ≥8 has demonstrated validity in the cancer setting,(32) a depression cutoff of ≥10 was used in this study since it is more commonly used among various populations and found to have a sensitivity of 0.77 (0.71 to 0.84) and a specificity of .94 (0.90 to 0.97).(32)
Anxiety
Anxiety severity and screening criteria were measured using the Generalized Anxiety Disorder-7 (GAD-7). The GAD-7 is a seven-item measure that scores each question from 0 to 3 for a range of scores from 0 to 21. Higher scores indicate greater anxiety.(33) The GAD-7 is employed as a screening tool for generalized anxiety disorder among cancer patients.(34) The internal consistency of the measure is excellent (Cronbach’s α=.92) with a good test-retest reliability (intra-class correlation = 0.83).(33) Diagnostic accuracy was assessed against another anxiety screen in patients with cancer and found to have an identical AUC of .81 (95% CI: .79-.82).(34) While various cut points have been used to identify cases of anxiety, a cutoff of ≥10 was used in this study corresponding to moderate anxiety.(33) A cutoff score of ≥10 has demonstrated a sensitivity of 89% and specificity of 82%.(33)
Statistical Analysis
Descriptive analyses were performed to assess distributions and central tendency for demographic and clinical covariates. Differences between patients with metastatic and localized cancer (or in remission) were assessed using independent samples t-tests and chi square tests. Bivariate correlations between psychological outcomes (depression and anxiety) and inflammatory markers (albumin, ANC, ALC, and NLR) were assessed using Spearman correlations to account for non-normally distributed variables. Linear regression models of depression and anxiety were created using albumin or NRL as the independent variable and controlling for age and sex. Exploratory analyses were performed using median split to assess the relationship between depression or anxiety and inflammatory markers (albumin and NLR) as categorical variables. Additional exploratory analyses were performed using descriptive statistics (distributions and central tendency) and tests of difference (independent samples t test and chi square) for patients who were taking antidepressant medications prior to the psychiatric visit. Statistical procedures were performed using the SPSS version 24 software (SPSS, Chicago, IL 2013), and all statistical tests were two-tailed with a .05 significance level.
Results:
Cohort Characteristics
The average age was 57.2 (SD 13.2) years old and 65% were female; 62% met depression screening criteria (PHQ-9≥10); 60% met anxiety screening criteria (GAD-7≥10); and 47% met screening criteria for both (Table 1). Seventy-six patients (76%) had solid tumor cancers and 21 patients (21%) had liquid tumors. Of the solid tumors, the most common were breast (n=20), thoracic or head and neck cancers (n=20), gastrointestinal cancers (n=12), melanoma or sarcoma (n=7), prostate cancer (n=6), genitourinary cancers (n=4), brain or central nervous system (CNS) tumors (n=4), and gynecologic cancers (n=3). Of the liquid tumors, the most common were acute leukemia (n=8), lymphoma (n=7), multiple myeloma (n=3), and chronic leukemia (n=2). Most patients (69%) were taking psychotropic medication prior to the initial psychiatry appointment. Prior to the visit, patients were most likely taking antidepressants (n=38), benzodiazepines (n=31), antipsychotics (n=8), anticonvulsants (n=5), or stimulants (n=2). Twenty patients presented on a combination of medications, most commonly an antidepressant and benzodiazepine. Of the antidepressant medications that patients were taking on presentation to clinic, SSRIs were the most common (n=21), followed by SNRIs (n=8), bupropion (n=7), mirtazapine (n=5), and ketamine (n=1). Antidepressant medications were started, changed, or increased at 45% of the psychiatry visits.
Table 1:
Clinical and demographic characteristics of the cohort
| Total (n=97) | Metastatic (n=51) | Localized cancer (n=46) | Difference | |
|---|---|---|---|---|
| M (SD) | M (SD) | M (SD) | ||
| Age (years) | 57.21 (13.2) | 57.6 (13.3) | 56.84 (13.2) | t=−.276, p=.78 |
| Anxiety (GAD-7) | 11.18 (6.3) | 10.82 (6.7) | 11.56 (5.9) | t=.584, p=.56 |
| Depression (PHQ-9) | 12.17 (6.6) | 11.47 (6.6) | 12.9 (6.7) | t=1.075, p=.29 |
| Time from blood draw (days) | 22.4 (37.2) | 16.3 (23.3) | 30.3 (48.8) | t=1.669, p=.10 |
| Albumin | 3.94 (.43) | 3.82 (.44) | 4.10 (.36) | t=3.180, p=.002 |
| ANC | 4.92 (4.6) | 5.28 (5.65) | 4.42 (2.9) | t=−.857, p=.394 |
| ALC | 1.48 (1.1) | 1.64 (1.30) | 1.26 (.64) | t=−1.823, p=.07 |
| NLR | 5.84 (8.7) | 5.00 (5.76) | 6.98 (11.6) | t=1.058, p=.29 |
| N (%) | N (%) | N (%) | ||
| Gender | ||||
| • Male | 35 (35%) | 18 (35%) | 16 (35%) | χ2=.004, p=.95 |
| • Female | 65 (65%) | 33 (64%) | 30 (65%) | |
| Disease Type | ||||
| • Solid Tumors | 76 (78%) | 34 (68%) | 41 (89%) | χ2=10.622, p=.001 |
| • Heme malignancy | 21 (22%) | 16 (32%) | 5 (11%) | |
| Time of Blood Draw | ||||
| Before appt | 63 (63%) | 34 (67%) | 28 (61%) | χ2=.358, p=.55 |
| After appt | 37 (37%) | 17 (33%) | 18 (39%) | |
| Blood draw within a week? | ||||
| • Yes | 45 (45%) | 27 (53%) | 17 (37%) | χ2= .266, p=.61 |
| • No | 55 (55%) | 24 (47%) | 29 (63%) | |
| Meets Depression Screening (PHQ-9D≥10) | ||||
| • Yes | 62 (62%) | 29 (57%) | 31 (67%) | χ2 =1.01, p=.32 |
| • No | 38 (38%) | 22 (43%) | 15 (33%) | |
| Meets Anxiety Screening (GAD-7≥10) | ||||
| • Yes | 60 (60%) | 31 (61%) | 27 (59%) | χ2=.027, p=.87 |
| • No | 40 (40%) | 20 (39%) | 19 (41%) | |
| Taking Psychotropic (prior to initial visit) | ||||
| • Yes | 69 (69%) | 32 (65%) | 35 (76%) | χ2= 1.904, p=.17 |
| • No | 31 (31%) | 19 (35%) | 11 (24%) | |
| Taking Antidepressant (prior to initial visit) | ||||
| • Yes | 38 (38%) | 18 (64%) | 19 (41%) | χ2 = .324, p=.57 |
| • No | 62 (62%) | 33 (35%) | 27 (59%) | |
| Plan to Start or Increase Antidepressant | ||||
| • Yes | 45 (45%) | 23 (45%) | 21 (45%) | χ2<.001, p=.98 |
| • No | 55 (55%) | 28 (55%) | 25 (55%) | |
Ninety-three out of 97 patients (96%) demonstrated values for albumin and the NLR obtained from a chemistry panel or complete blood count, respectively. No other markers of inflammation (e.g., CRP or Erythrocyte Sedimentation Rate) were found consistently in the sample. Almost half (45%) had blood draws within a week of their initial psychiatric appointment and most blood draws were done prior to the visit (63%). The average length of time between the blood draw and the initial visit was 22 days (SD 37 days). The average serum albumin concentration was 3.94 g/dL (SD 0.43g/dL) but it was greater in patients with localized cancer and lower in patients with metastatic cancer (4.10 g/dL versus 3.82 g/dL, respectively) (p=.002). The mean NLR was 5.8 (SD 8.7) and there was no difference between patients with localized or metastatic cancer.
Half (51%) of the sample had metastatic disease (Table 1). The only statistically significant differences between patients with metastatic and localized disease were lower albumin levels in patients with metastatic disease (3.83 versus 4.10, p=.002) and fewer patients with hematologic malignancy had localized disease or were in survivorship (χ2= 10.622, p=.001). Of note, the time from blood draw was shorter in the metastatic group (16 days versus 30 days, p=.10) and the ALC was higher in the metastatic group (1.64 versus 1.26, p=.07) but neither was statistically significant.
Correlations between Inflammatory Markers and Anxiety or Depression
Correlation matrices highlight the relationships between timing of blood draws (i.e., recorded at any point in the record versus within seven days) and between patient types (i.e., metastatic cancer versus localized cancer).
Albumin or NLR drawn at any time point for patients with localized or metastatic cancer were not correlated with anxiety or depression (Table 2A). In patients with metastatic cancer only (n=51), albumin drawn at any time point was correlated with anxiety (r=−.28, p=.04) and depression (r=−.40, p=.004). Albumin drawn within seven days of the psychiatric appointment (n=27) was more strongly correlated with anxiety (r=−.40, p=.02) and depression (r=−.64, p<.001) (Table 2B). There were no significant correlations between NLR and anxiety or depression in patients with metastatic cancers at either time point.
Table 2:
Correlation matrices for A) all patients with labs drawn at any time point (n=93) or within seven days of psychiatric appointment (n=46) and B) for patients with metastatic cancer at any time point (n=51) or within seven days of psychiatric appointment (n=27).
| A) All Patients | ||||||||
|---|---|---|---|---|---|---|---|---|
| Labs drawn at any time point (n=93) | ||||||||
| Age | Dep | Anx | Albumin | ANC | ALC | NLR | ||
| Labs drawn within 7 days (n=46) | Age | −.16 | −.22 | −.09 | .16 | .02 | .13 | |
| Dep | −.23 | .63*** | −.09 | −.15 | .05 | −.13 | ||
| Anx | −.19 | .55*** | −.07 | −.07 | .08 | −.08 | ||
| Albumin | −.04 | −.17 | −.01 | .02 | .15 | −.11 | ||
| ANC | .09 | .20 | .25 | .08 | −.00 | .62*** | ||
| ALC | −.17 | −.04 | .09 | .08 | .03 | −.71*** | ||
| NLR | .21 | −.04 | .02 | −.02 | .62*** | −.69*** | ||
| B) Patients with Metastatic Cancer | ||||||||
|---|---|---|---|---|---|---|---|---|
| Labs drawn at any time point (n=51) | ||||||||
| Age | Dep | Anx | Albumin | ANC | ALC | NLR | ||
| Labs drawn within 7 days (n=27) | Age | −.13 | −.26 | −.04 | .19 | −.15 | .22 | |
| Dep | −.20 | .72*** | −.40** | .23 | −.02 | −.02 | ||
| Anx | −.33 | .59** | −.28* | .10 | .10 | −.03 | ||
| Albumin | .02 | −.64*** | −.40* | −.12 | .17 | −.19 | ||
| ANC | .18 | .30 | .16 | −.01 | −.02 | .66*** | ||
| ALC | −.44* | −.05 | .06 | .14 | .10 | −.70*** | ||
| NLR | .36 | .04 | .15 | −.11 | .72*** | −.55** | ||
Note:
p<.05
p<.01
p<.001. ALC, absolute lymphocyte count; ANC, absolute neutrophil count; Anx, anxiety; Dep, depression; NLR, neutrophil to lymphocyte ratio
Linear Regression Models
Linear regression modeling of patients with metastatic disease found models of albumin taken from any time point and depression or anxiety (Table 3, Models 1 and 2) were not statistically significant when controlling for age and sex (F=1.841, p=.16; F=2.210, p=.10 respectively). However, the model for depression severity based on albumin (Table 3, Model 3) was statistically significant when selecting for patients whose labs were drawn within seven days (β=−.63, p=.002) and accounted for 32% of the depression severity variance. The model for anxiety severity based on albumin (Table 3, Model 4) trended towards significance (F=2.527, p=.08).
Table 3:
Association between albumin and depression or anxiety in patients with metastatic cancers (n=50) using all available laboratory values (Model 1) and those drawn within 1 week of visit in CL clinic (Model 2) (n=27).
| Depression | Confidence Interval | ||||
|---|---|---|---|---|---|
| Model 1 (any available lab value) | Regression Coefficient | β | P | Lower | Upper |
| Age | −.03 (.07) | −.06 | .65 | −.166 | .105 |
| Sex | −1.21 (1.89) | −.09 | .53 | −5.001 | 2.587 |
| Albumin | −4.30 (2.06) | −.29 | .04* | −8.450 | −.158 |
| adjusted R2 .05; F=1.841, p=.15 | |||||
| Anxiety | Confidence Interval | ||||
| Model 1 (any available lab value) | Regression Coefficient | β | P | Lower | Upper |
| Age | −.09 (.07) | −.18 | .21 | −.225 | .050 |
| Sex | −.25 (1.91) | −.02 | .90 | −4.094 | 3.589 |
| Albumin | −4.46 (2.09) | −.30 | .04* | −8.660 | −.264 |
| adjusted R2 .069 F=2.210, p=.10 | |||||
| Depression | Confidence Interval | ||||
| Model 2: (lab within a week) | Regression Coefficient | β | P | Lower | Upper |
| Age | −.13 (.10) | −.22 | .20 | −.331 | .073 |
| Sex | 1.60 (2.17) | .13 | .47 | −2.892 | 6.094 |
| Albumin | −9.56 (2.69) | −.63 | .002** | −15.141 | −3.975 |
| adjusted R2 .32, F=4.927, p=.009** | |||||
| Anxiety | Confidence Interval | ||||
| Model 2: (lab within a week) | Regression Coefficient | β | P | Lower | Upper |
| Age | −.22 (.12) | −.36 | .08 | −.467 | .024 |
| Sex | −.42 (2.64) | −.03 | .87 | −5.890 | 5.044 |
| Albumin | −6.03 (3.28) | −.36 | .08 | −12.818 | .769 |
| adjusted R2 .16, F=2.527, p=.08 | |||||
Note:
p<.05
p<.01.
Linear regression modeling of NLR at any time point or within one week of psychiatric visit was not associated with anxiety (F=0.70, p=.56; F=1.22, p=.30) or depression (F=0.51, p=.67; F= 1.15, p=.33) in patients with metastatic cancers.
Association between Albumin and Depression or Anxiety Screening Criteria
A median split for albumin in patients with metastatic cancer was found at 3.8g/dL. Therefore, low albumin was defined as 3.8g/dL or below and high albumin (normal albumin) was defined as above 3.8g/dL. Meeting screening criteria for either depression or anxiety was associated with having low albumin (χ2= 11.055, p=.001; χ2= 4.433, p=.04, respectively) (Table 4).
Table 4:
Patients with metastatic disease (n=51)
| Depression Criteria (PHQ-9≥10) | Chi Square 11.055, p=.001** | ||
|---|---|---|---|
| Albumin Median Split | No | Yes | Total |
| Above 3.8 g/dL | 15 | 6 | 21 |
| 3.8 g/dL or below | 7 | 23 | 30 |
| Total | 22 | 29 | 51 |
| Anxiety Criteria (GAD-7≥10) | Chi Square 4.433, p=.04* | ||
|---|---|---|---|
| Albumin Median Split | No | Yes | Total |
| Above 3.8 | 12 | 9 | 21 |
| 3.8 or below | 8 | 22 | 29 |
| Total | 20 | 31 | 51 |
Note:
p<.05
p<.01. GAD-7, Generalized Anxiety Disorder-7; PHQ-9, Patient Health Questionnaire-9.
Inflammatory markers and antidepressant treatment resistance: exploratory analysis of patients taking an antidepressant prior their psychiatric visit (n=38)
Albumin and NLR were assessed along with depression and anxiety in patients who were already taking an antidepressant prior to their psychiatric evaluation to explore whether higher inflammation was associated with less antidepressant response (i.e., higher depression scores). Overall, these patients had lower albumin concentrations than patients who were not yet taking an antidepressant (albumin concentration of 3.89 g/mL [SD .37] versus 3.99g/mL [SD .46] respectively) (t=.56, p=.28). Patients on antidepressants also had higher depression levels than patients who were not taking an antidepressant prior to the psychiatric visit (12.7 versus 11.7 respectively).
Among patients on antidepressants prior to the visit (n=38), those with low albumin (defined by median split) had slightly higher depression scores than patients with high albumin (12.0 versus 11.7 respectively) (t=−0.12, p=.54). Anxiety scores between the high/low albumin groups were also not significantly different (8.7 versus 8.6, respectively) (t=0.02), p=.98). Patients on antidepressants had lower NLR values than patients not on antidepressants, 5.2 (SD 7.6) versus 6.4 (SD 9.6) (t=1.36, p=.17). For patients already taking antidepressants, higher NLR (i.e., greater inflammation) was linked to greater depression (13.7 versus 10.2) and anxiety (10.5 versus 7.3) although the values were not significant (t=1.75 p=.09) and (t=1.52, p=.14) respectively.
Discussion:
This study found that routine laboratory values drawn outside of the psychiatric appointment were available and informative for the management of anxiety and depression in patients with cancer who were evaluated at a cancer center. All major cancer types were represented along with a variety of patients in remission or receiving active treatment for cancer. Although this was a single center study, the population is likely representative of other psychiatric oncology clinics where patients are being treated for cancer at the same institution. Patients were expectedly symptomatic with high levels of anxiety and depressive symptoms as predicted by the referral population. The symptom severity is also reflected by the management plans to start or titrate up antidepressant medications at almost half of these initial appointments.
This study is unique in that it provides information about symptom severity in cancer patients who are referred to psychiatry as opposed to the affective symptoms of a general oncology population (i.e., most anxiety or depression prevalence studies in cancer populations). These findings reinforce the need to move beyond a trial-and-error approach for highly symptomatic patients who do not have the time or quality of life to tolerate multiple lines of antidepressant or psychotherapeutic treatments.
Of the two routine laboratory values, only albumin demonstrated a clear trend in its association with anxiety and depression. NLR was not associated with anxiety or depression in any of the analyses. Although the exploratory analysis of patients already on antidepressants found that higher NLR was linked to higher depression and anxiety scores, these values were not statistically significant but may warrant further evaluation as a predictive marker in this setting. There are many reasons why neutrophil and lymphocyte concentrations change during the cancer trajectory (e.g., anti-neoplastic therapies, infections, bone marrow failure syndromes) which may preclude finding a relationship with its variability in the cancer setting despite its association with depression and other affective symptoms in non-cancer studies.
Albumin is an indirect measure of general innate inflammation and decreases in the presence of systemic inflammation. Several studies have demonstrated a relationship between depression and albumin while controlling for markers of nutritional status. In general, pre-albumin is a better marker of nutrition than albumin. As a negative acute phase reactant, albumin was associated with approximately 32% variation in depression scores, which is similar to the association between CRP and depression.(25, 35) The range over which albumin changes with depression is much smaller than its positive acute phase reactant counterpart, CRP, which changes exponentially.(31) This could make the interpretability of albumin less ideal even though low albumin based on a median split was associated with meeting screening criteria for both anxiety and depression. Also, rare conditions such as protein losing nephropathies and enteropathies will cause hypoalbuminemia regardless of systemic inflammation and would probably preclude albumin as a biomarker for depression or anxiety in those scenarios.
Timing of the lab draw, and cancer state (localized versus metastatic) mattered. The relationship between albumin and anxiety or depression was significant only for patients with metastatic disease and was strengthened when labs were within seven days of the appointment. Differences in albumin drawn within seven days predicted depression scores when controlling for age and sex. The half-life of albumin is 20 days so weeks are required to adjust to changes in inflammation.(31) Although inflammation can change acutely in response to infection or another factor, anxiety and depression symptoms generally change in response to chronic inflammation, especially when the source of inflammation is ongoing as in patients with metastatic cancer. Patients with localized cancers may have developed chronic inflammation due to receipt of anti-neoplastic therapy, radiation, or surgery but these additional reasons for chronic inflammation and the variable timing in follow up precluded any relationship between inflammation and anxiety or depression symptoms in patients with localized cancer. In other words, this patient group was heterogeneous, and a relationship may emerge when that is taken into consideration.
Inflammation has been associated with depression treatment refractoriness and as a predictive biomarker to antidepressant response.(21, 29) An antidepressant treatment that addresses inflammation may be particularly efficacious in patients with cancer given the prevalence of inflammation in cancer settings. Over two-thirds of patients (69%) were already taking psychotropic medications, mostly antidepressants, which may have limited efficacy in patients with cancer and high levels of inflammation. Additionally, medications were added and titrated in almost half (45%) of the psychiatric appointments.
Cancer-related inflammation is a direct byproduct of having cancer through its genomic instability and other cancer-tumor environment interactions and the related concept of cancer-related immunity is now used to direct anticancer treatments (i.e., checkpoint inhibitor therapies).(16) Systemic inflammation tends to deplete mono-amine neurotransmitters in the central nervous system.(36) Both serotonin and dopamine are depleted, as shown by fMRI data, but only serotonin is restored with typical selective serotonin reuptake inhibitors (SSRIs).(36) The only drug that is approved for the treatment of depression and upregulates dopamine is bupropion, which has a long history of being used in combination with SSRIs in patients who are not responding to monotherapy.(37) The underlying physiology of systemic inflammation and its downstream effects on the CNS (i.e., depletion of dopamine) appear to explain why the addition of bupropion would be effective (i.e., increased availability of dopamine).(38) In addition, treatments that lower inflammation (e.g., anti-inflammatory medications that block COX1/2 pathways, monoclonal antibody medication that antagonize cytokine receptors [IL-6]) are associated with decreased rates of depression and depressive symptoms.(39)
Study Limitations
This study was limited by its cross-sectional, retrospective design at a single institution and a high level of cancer-related variability (e.g., different cancer types, stages, treatment types). These data are taken from initial appointments only and longitudinal data are not available. These data are hypothesis-generating and descriptive. Therefore, potential mediating variables (e.g., cancer type and stage) were not incorporated into the statistical analyses. For example, cancer type and its influence on inflammation (e.g., albumin) and depression variability should be assessed in future studies. Cancer subtypes may be associated with depression irrespective of inflammation and future studies of cancer subtypes should account for this variability. In addition, this study did not capture the role of other important oncologic variables such as performance status and tumor burden, which may be associated with depression independently of inflammation.
Clinical Implications
This study provides a survey of patients with cancer seen in a dedicated psychiatric clinic, their symptom severity, anxiety and depression treatment plans, and available labs from which to draw conclusions about their inflammatory status. The study describes a psychiatric need and an opportunity to potentially incorporate inflammation into the care of affective disorders in patients with cancer addressing this an unmet need. Since inflammation is relevant for affective states in many chronic medical illnesses, these findings may extrapolate to other settings. Although the relationship between depression and inflammation exists in medically healthy patients, the relationship is especially salient for patients with cancer for whom inflammation is also a predictor of mortality and is associated with other adverse quality of life symptoms such as fatigue, poor sleep, and pain.(40)
The available information provides a general picture of their inflammatory states which could be used to either hypothesize treatment refractoriness or predict treatment responsiveness in depression treatments that address inflammation (e.g., lowering inflammation and/or increasing available monoamine neurotransmitters). Both hypotheses are important for patients with cancer who exhibit high levels of symptom severity without targeted treatments for these symptoms. These findings warrant future studies to confirm the results. In addition, future studies may distinguish to what extent depression is related to inflammation (e.g., hypoalbuminemia) versus progressive or worsening disease states. The ability for inflammation to help guide antidepressant treatment may be tremendously relevant for the treatment of anxiety and depression in patients with cancer.
Conclusion
In summary, psycho-oncology clinics have reasonable access to laboratory assessments drawn during oncology visits. Albumin obtained within a week may provide a useable assessment of inflammatory status and predict depression and treatment refractoriness. The incorporation of inflammatory markers may help inform symptom management by identifying patients who may have recalcitrant symptom or benefit from an approach that address ongoing chronic inflammation. These findings warrant further exploration in these areas and with additional markers of inflammation. The type of inflammatory marker (albumin versus NLR), the setting (localized versus metastatic), and timing of blood draw (at any point versus within 1 week) are all important for understanding how to apply this well-established relationship for clinical benefit to address these highly prevalent symptoms.
Acknowledgements:
Funding: Not applicable.
Ethical Approval Statement:
The study was approved by the Memorial Sloan Kettering Cancer Center Institutional Review Board (IRB) on 12/10/2018 as protocol number 18–526 (Routine inflammatory markers among mental health clinic patients being treated for depression).
Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Footnotes
Conflict of Interests: No conflict of interest reported by authors.
References
- 1.Watson M, Dunn J. The multidisciplinary art and science of cancer care: integrating psycho-oncology. Future Oncol. 2016;12(24):2775–8. [DOI] [PubMed] [Google Scholar]
- 2.Mitchell AJ, Chan M, Bhatti H, Halton M, Grassi L, Johansen C, et al. Prevalence of depression, anxiety, and adjustment disorder in oncological, haematological, and palliative-care settings: a meta-analysis of 94 interview-based studies. Lancet Oncol. 2011;12(2):160–74. [DOI] [PubMed] [Google Scholar]
- 3.Pinquart M, Duberstein PR. Depression and cancer mortality: a meta-analysis. Psychol Med. 2010;40(11):1797–810. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Arrieta O, Angulo LP, Nunez-Valencia C, Dorantes-Gallareta Y, Macedo EO, Martinez-Lopez D, et al. Association of depression and anxiety on quality of life, treatment adherence, and prognosis in patients with advanced non-small cell lung cancer. Ann Surg Oncol. 2013;20(6):1941–8. [DOI] [PubMed] [Google Scholar]
- 5.Pfeiffer PN, Ganoczy D, Ilgen M, Zivin K, Valenstein M. Comorbid anxiety as a suicide risk factor among depressed veterans. Depress Anxiety. 2009;26(8):752–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Melton L, Krause D, Sugalski J. Psychology Staffing at Cancer Centers: Data From National Comprehensive Cancer Network Member Institutions. JCO Oncol Pract. 2020;16(11):e1343–e54. [DOI] [PubMed] [Google Scholar]
- 7.Young K, Singh G. Biological Mechanisms of Cancer-Induced Depression. Front Psychiatry. 2018;9:299. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sotelo JL, Musselman D, Nemeroff C. The biology of depression in cancer and the relationship between depression and cancer progression. Int Rev Psychiatry. 2014;26(1):16–30. [DOI] [PubMed] [Google Scholar]
- 9.Walker J, Sawhney A, Hansen CH, Ahmed S, Martin P, Symeonides S, et al. Treatment of depression in adults with cancer: a systematic review of randomized controlled trials. Psychol Med. 2014;44(5):897–907. [DOI] [PubMed] [Google Scholar]
- 10.Ioannidis JP. Effectiveness of antidepressants: an evidence myth constructed from a thousand randomized trials? Philos Ethics Humanit Med. 2008;3:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Rush AJ, Trivedi MH, Wisniewski SR, Nierenberg AA, Stewart JW, Warden D, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry. 2006;163(11):1905–17. [DOI] [PubMed] [Google Scholar]
- 12.Raison CL, Capuron L, Miller AH. Cytokines sing the blues: inflammation and the pathogenesis of depression. Trends Immunol. 2006;27(1):24–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Maes M, Berk M, Goehler L, Song C, Anderson G, Galecki P, et al. Depression and sickness behavior are Janus-faced responses to shared inflammatory pathways. BMC Med. 2012;10:66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Costello H, Gould RL, Abrol E, Howard R. Systematic review and meta-analysis of the association between peripheral inflammatory cytokines and generalised anxiety disorder. BMJ Open. 2019;9(7):e027925. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Vogelzangs N, Beekman AT, de Jonge P, Penninx BW. Anxiety disorders and inflammation in a large adult cohort. Transl Psychiatry. 2013;3:e249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Colotta F, Allavena P, Sica A, Garlanda C, Mantovani A. Cancer-related inflammation, the seventh hallmark of cancer: links to genetic instability. Carcinogenesis. 2009;30(7):1073–81. [DOI] [PubMed] [Google Scholar]
- 17.Basterzi AD, Aydemir C, Kisa C, Aksaray S, Tuzer V, Yazici K, et al. IL-6 levels decrease with SSRI treatment in patients with major depression. Hum Psychopharmacol. 2005;20(7):473–6. [DOI] [PubMed] [Google Scholar]
- 18.Tynan RJ, Weidenhofer J, Hinwood M, Cairns MJ, Day TA, Walker FR. A comparative examination of the anti-inflammatory effects of SSRI and SNRI antidepressants on LPS stimulated microglia. Brain Behav Immun. 2012;26(3):469–79. [DOI] [PubMed] [Google Scholar]
- 19.Howren MB, Lamkin DM, Suls J. Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis. Psychosom Med. 2009;71(2):171–86. [DOI] [PubMed] [Google Scholar]
- 20.Strawbridge R, Hodsoll J, Powell TR, Hotopf M, Hatch SL, Breen G, et al. Inflammatory profiles of severe treatment-resistant depression. J Affect Disord. 2019;246:42–51. [DOI] [PubMed] [Google Scholar]
- 21.Haroon E, Daguanno AW, Woolwine BJ, Goldsmith DR, Baer WM, Wommack EC, et al. Antidepressant treatment resistance is associated with increased inflammatory markers in patients with major depressive disorder. Psychoneuroendocrinology. 2018;95:43–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Huang TL. Lower serum albumin levels in patients with mood disorders. Chang Gung Med J. 2002;25(8):509–13. [PubMed] [Google Scholar]
- 23.Demir S, Atli A, Bulut M, Ibiloglu AO, Gunes M, Kaya MC, et al. Neutrophil-lymphocyte ratio in patients with major depressive disorder undergoing no pharmacological therapy. Neuropsychiatr Dis Treat. 2015;11:2253–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Xu L, Pan Q, Lin R. Prevalence rate and influencing factors of preoperative anxiety and depression in gastric cancer patients in China: Preliminary study. J Int Med Res. 2016;44(2):377–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.McFarland DC, Breitbart W, Miller AH, Nelson C. Depression and Inflammation in Patients With Lung Cancer: A Comparative Analysis of Acute Phase Reactant Inflammatory Markers. Psychosomatics. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Keller U. Nutritional Laboratory Markers in Malnutrition. J Clin Med. 2019;8(6). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Giannousi Z, Gioulbasanis I, Pallis AG, Xyrafas A, Dalliani D, Kalbakis K, et al. Nutritional status, acute phase response and depression in metastatic lung cancer patients: correlations and association prognosis. Support Care Cancer. 2012;20(8):1823–9. [DOI] [PubMed] [Google Scholar]
- 28.Pascoe MC, Skoog I, Blomstrand C, Linden T. Albumin and depression in elderly stroke survivors: An observational cohort study. Psychiatry Res. 2015;230(2):658–63. [DOI] [PubMed] [Google Scholar]
- 29.Harley J, Luty S, Carter J, Mulder R, Joyce P. Elevated C-reactive protein in depression: a predictor of good long-term outcome with antidepressants and poor outcome with psychotherapy. J Psychopharmacol. 2010;24(4):625–6. [DOI] [PubMed] [Google Scholar]
- 30.Pineiro M, Pato R, Soler L, Pena R, Garcia N, Torrente C, et al. A new automated turbidimetric immunoassay for the measurement of canine C-reactive protein. Vet Clin Pathol. 2018;47(1):130–7. [DOI] [PubMed] [Google Scholar]
- 31.Fanali G, di Masi A, Trezza V, Marino M, Fasano M, Ascenzi P. Human serum albumin: from bench to bedside. Mol Aspects Med. 2012;33(3):209–90. [DOI] [PubMed] [Google Scholar]
- 32.Thekkumpurath P, Walker J, Butcher I, Hodges L, Kleiboer A, O’Connor M, et al. Screening for major depression in cancer outpatients: the diagnostic accuracy of the 9-item patient health questionnaire. Cancer. 2011;117(1):218–27. [DOI] [PubMed] [Google Scholar]
- 33.Spitzer RL, Kroenke K, Williams JB, Lowe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Archives of internal medicine. 2006;166(10):1092–7. [DOI] [PubMed] [Google Scholar]
- 34.Esser P, Hartung TJ, Friedrich M, Johansen C, Wittchen HU, Faller H, et al. The Generalized Anxiety Disorder Screener (GAD-7) and the anxiety module of the Hospital and Depression Scale (HADS-A) as screening tools for generalized anxiety disorder among cancer patients. Psychooncology. 2018;27(6):1509–16. [DOI] [PubMed] [Google Scholar]
- 35.Wium-Andersen MK, Orsted DD, Nielsen SF, Nordestgaard BG. Elevated C-reactive protein levels, psychological distress, and depression in 73, 131 individuals. JAMA Psychiatry. 2013;70(2):176–84. [DOI] [PubMed] [Google Scholar]
- 36.Felger JC, Lotrich FE. Inflammatory cytokines in depression: neurobiological mechanisms and therapeutic implications. Neuroscience. 2013;246:199–229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Fava M, Rush AJ, Thase ME, Clayton A, Stahl SM, Pradko JF, et al. 15 years of clinical experience with bupropion HCl: from bupropion to bupropion SR to bupropion XL. Prim Care Companion J Clin Psychiatry. 2005;7(3):106–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Felger JC. The Role of Dopamine in Inflammation-Associated Depression: Mechanisms and Therapeutic Implications. Curr Top Behav Neurosci. 2017;31:199–219. [DOI] [PubMed] [Google Scholar]
- 39.Kohler O, Benros ME, Nordentoft M, Farkouh ME, Iyengar RL, Mors O, et al. Effect of anti-inflammatory treatment on depression, depressive symptoms, and adverse effects: a systematic review and meta-analysis of randomized clinical trials. JAMA Psychiatry. 2014;71(12):1381–91. [DOI] [PubMed] [Google Scholar]
- 40.Bower JE, Ganz PA, Irwin MR, Kwan L, Breen EC, Cole SW. Inflammation and behavioral symptoms after breast cancer treatment: do fatigue, depression, and sleep disturbance share a common underlying mechanism? J Clin Oncol. 2011;29(26):3517–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.